Distributed Alamouti Relay Beamforming Scheme in Multiuser Relay Networks

We design distributed relay beamforming in a multiuser peer-to-peer relay network. By exploring Alamouti code at both sources and relays, we propose a rank-two Alamouti-based distributed relay beamforming scheme to minimize per relay power, while meeting the signal-to-interference-and-noise ratio targets. For the nonconvex optimization problem, we propose a rank-constrained separable semidefinite relaxation approach to find an approximate solution, and provide conditions for which it produces an optimal solution and a bound on the gap to the optimal performance. Compared with the traditional rank-one distributed relay beamforming scheme, our proposed Alamouti-based rank-two distributed relay beamforming offers a significantly higher likelihood to produce an optimal solution and a better capability to maintain small performance degradation as the network size increases. As a result, it provides substantially improved relay power efficiency.